The purpose of a compressor is to even out the audio signal in order to tame loud transients and bring up lower amplitude material. Compressors were born out of necessity to fill that very straightforward role, to protect radio transmitters from overly hot signals, and to keep recordings in a narrower dynamic range that was easier to reproduce on tape or record.
But the character of a compressor began to be exploited back in the 1960s to become part of the sonic palette. As studios became tools for the creation of sound and not just the capturing of it, so too did compressors become tools for creating a sonic signature, not just managing levels. How exactly did they do that? By utilizing a compressor’s functions to interact with the program material in a creative way. Let’s look at those functions, which all compressors use (but not all have controls for), in the order in which they occur.
The Difference Between Compressors and Limiters
A common question we get is "What's the difference between a compressor and a limiter?" It really all comes down to the compression ratio that is used. A limiter's threshold is to the limit for the maximum level and often utilizes higher ratios above 10:1. Compressors employ lower ratios (think anything below 5:1) and often have faster attack and release times.
The Functions Of A Compressor/Limiter
The threshold setting determines the point at which the unit begins working. Once the signal level has crossed the threshold, it is then reduced by the set ratio.
The amount of total gain reduction applied to the signal is determined by the ratio. Using a setting of 2:1 as an example, for every 2db over the threshold the signal goes, its output is only 1db. So if the signal is 6db over the threshold, the compressor will reduce the signal by half or 3db. Ratios of 10:1 and higher are usually considered limiting. Most compressors can handle this ratio, so they are often referred to as ‘compressor/limiters.’ Broadcasters use something called “brick wall” limiting to ensure that the signal will not exceed a set level. This ratio is very high, with some limiters having settings of 100:1 all the way to infinity to 1.
Attack refers to the time it takes for the unit to begin working once the signal crosses the threshold. For most modern units the attack time is adjustable, while in many vintage examples the attack time is fixed. Correctly setting the attack for signals containing transients is critical. Not leaving enough time for the transient information to sneak through before the unit clamps down can seriously degrade signal quality. Attack times between 20 and 800 μs (microseconds) would be considered fast, while slower times range from 10 to 100 ms (milliseconds). A millisecond equals one thousand microseconds.
HARD KNEE / SOFT KNEE
Hard Knee/Soft Knee refers to the unit’s ability to either gradually apply gain reduction (soft knee) or make it quicker (hard knee) and therefore more apparent. A hard knee may be desirable if a limiter is used to control peaks, while a soft knee may be more appropriate for a compressor set to squeeze a signal and get a more musical effect. The word “knee” comes from the shape of the slope when the response curve is shown on a graph. Some compressors offer a choice of a hard or soft knee, or even a variable between those extremes.
The amount of time the unit takes to stop compressing and return to unity gain is referred to as release time. Just like attack time, the release time is usually adjustable on modern units, but often fixed on vintage versions. Release times generally overlap the longer attack times, often ranging anywhere from 40 milliseconds up to 5 seconds Careful adjustment is important here as well, due to an effect known as “pumping.” When release times are set inappropriately to the program material, a rapid raising and lowering of the signal is heard while the unit tries to compensate for the mismatch. This audible artifact is to be avoided unless it’s being done on purpose.
The dynamic range of a signal is the distance in decibels from its lowest level to its highest or peak level. Compressors (and limiters) work to narrow this range, i.e. to literally compress it. Though a compressor only reduces the peak of a signal, this in turn increases the audibility of the quieter parts, thus bringing the louder and quieter limits of the signal closer together.
After being compressed, the signal is by definition at a lower amplitude than when it entered the unit. To return the overall level to the original amplitude requires using makeup gain. This does not undo the effect of the compressor, but brings the compressed or limited signal back up to an optimal recording level. Certain compressor models will have a meter that can be set to Gain Reduction, or GR; this shows you the attenuation in dB so that you can apply the correct amount of makeup gain to compensate.
Different Types Of Compressors/Limiters
The earliest types of compressors, unsurprisingly, relied on vacuum tubes to achieve their gain reduction, as all audio gear in this era relied on tubes for power, amplification, and rectifying. Let’s begin there.
Delta-Mu or Variable-Mu® Compressors/Limiters
When you think of classic original compressors, the granddaddy of them all is the Fairchild, either the mono 660 model or its bigger brother, the stereo 670. Other important models are the Universal Audio 175-B and the RCA BA-6A. All of them do their thing in a program-dependent way, meaning that the ratio of gain reduction is increased as the signal gets hotter ( they don’t even have a ratio control knob). Technically, as the signal feeding the tube increases, the amount of current sent to the grid decreases, dropping the overall level, while increasing the compression ratio in a non-linear way, with a built-in soft knee and a relatively slow attack time. This results in a smooth compression that is rich in harmonics and directly related to the program content. This makes a Delta-Mu compressor an ideal unit for bus compression or when you're mastering and looking to do compression in broader strokes over a complex mix. One of the most sought-after modern units is the Manley Vari-Mu®.
Suggested Delta-Mu/Variable-Mu® Compressors/Limiters: Fairchild 660, Fairchild 670, UnderTone Audio UnFairchild 670M II, Universal Audio 175-B, Manley Labs Vari-Mu, Retro Instruments 175-B Vintage King Edition, Retro Instruments Sta-Level, Retro Instruments Doublewide, Retro Instruments Doublewide II, Retro Instruments Revolver, RCA BA-6A, Chandler Limited RS124, Highland Dynamics BG2 Compressor, IGS Audio Tubecore 3U, IGS Audio Tubecore 500, and Gyraf G22 Dual Stereo Vari-Mu Compressor.
The other early compressor form is the optical, or opto for short. Its name refers to the use of an optical element such as a light bulb or LED, that shines on a light-sensitive resistor, which in turn controls the gain reduction. Although light itself is fast, this type of compressor is still on the slow side, as the resistor itself is not fast-acting. It also cannot be as finely adjusted as a Delta-Mu type, so there are no attack or release controls on the earliest models. Instead, you have a resistor that reacts fairly quickly to latch onto the signal (about 10 ms) and also lets go quickly, but then tapers off while the resistor recovers (even up to 5 seconds), before going back to unity gain. This makes the whole compression cycle smoother and more musical. The standout early comp of this type is the Teletronix LA-2A, which does employ tubes but not for gain reduction, so it is not a Delta-Mu type, but an opto, using the legendary T-4 photoresistor cell to do its magic. The opto-compressors are generally better at program material with less transients (think strings and vocals), and since they have few controls they do not see much use for bus compression or mastering.
Field-Effect Transistor (FET) Compressors/Limiters
The FET was patented way back in the 1920s, but in the 1960s they began to be used as the variable resistor in a compressor. Using a FET for this purpose offered an alternative to the optical and Delta-Mu styles of the times, providing a very colored sound,the distinctive character of which played right into the growing use of the recording studio as a sonic tool. The most famous example of an FET compressor is the ubiquitous 1176 LN developed by Bill Putnam and Urei. Still made today, it is revered, cloned, and sought for its unique tone and aggressive character.
Diode Bridge Compressors/Limiters
Although not very common today, a couple of the most famous compressors in history were diode bridge designs: the legendary Neve 33609 and its ancestor, the 2254. Originally designed for use in Neve consoles, these have often been described as "desert island" compressors – very musical with pleasing harmonic distortion. The design allows for lots of control over the compression parameters, and they see plenty of use on the mix bus and in mastering. The 33609 is still in production from AMS Neve, a testament to its enduring popularity.
Voltage Controlled Amplifier (VCA) Compressors/Limiters
Technically almost all compressors are voltage controlled devices, but when we speak of VCA comps, it usually means that a transistor on an integrated circuit (IC) chip is involved. One of the earliest of this type is the vaunted dbx 160 VU from the 1970s, which is grainy,grungy, and a delight on bass and drums. Since the VCA works on the peak signal, it’s excellent at taming the transients. Generally, all the controls for the typical functions are found on a VCA: threshold, ratio, attack, release. The compression on modern designs is smooth and agile, as they offer more control over all the parameters (and a bit less character than an FET or opto). The VCA is the compressor of choice on the mix bus and in mastering.
Suggested VCA Compressors/Limiters: dbx 160VU, SSL G Series Console Bus Compressor, API 527, API 529, Overstayer Stereo Voltage Control, Vertigo VSC-2, Smart Research C2, Smart Research C1LA, API 2500, Shadow Hills Dual Vandergraph, Shadow Hills Mastering Compressor, Dramastic Obsidian Stereo Compressor, dbx 560A, WesAudio Dione 500 Series Stereo Bus and Dangerous Compressor.
Different Uses For Compressors/Limiters
This is a great technique for retaining an instrument’s natural sound while adding a bit of power and excitement to it. By splitting the signal, you keep one channel clean and insert a compressor on the second. Adjusting the comp for the desired effect, you bring the compressed signal up just under the clean signal. A subtle move, this technique does a good job of producing an open yet slightly more present sound.
CHAIN OR SERIAL COMPRESSION
Most often used for evening out vocals, chain compression links two or more units together in series. Like EQ, compression artifacts can be less audible at heavier settings when the job is split over more than one unit. In the vocal example, a ratio setting of 10:1 or higher (which is considered limiting at that point) would be a good start for unit 1. With a light threshold setting and fast attack, its job is to knock down any erroneous peaks in the signal. Unit 2 would have a much lower ratio setting (for a natural sound), say 2:1 or 3:1 and an appropriate threshold setting for the desired amount of even compression.
Multi-band compressors are essentially multiple units rolled into one, each dedicated to its own band of frequencies. Often used by mastering engineers, multi-band compressors are a handy tool for mixers when dealing with instruments with broad frequency ranges, as well as typically unruly signals such as bass guitar.
A favorite among mixers, this type of compression assigns a number of units, generally stereo, each to its own buss. Each unit is set with parameters complementing the others. In other words, unit 1 may be set up as a fast attack limiter, unit 2 with a fairly heavy 6:1 comp ratio, unit 3 with a gentle 2:1 ratio,and so on.
If you're interested in learning more or would like order any of the compressors/limiters mentioned in this Buyer's Guide, contact a Vintage King Audio Consultant via email or by phone at 866.644.0160.